Mammographic density is one of the strongest risk factors for breast cancer. Women with the highest mammographic density are at a four- to six-fold greater risk of breast cancer than women with the lowest density. Recently for other chronic diseases (e.g., coronary artery disease), we have seen proof-of-principle that utilizing a reliably measured, heritable quantitative trait (e.g., circulating lipids) that is a strong risk factor for the outcome can identify novel loci for the disease that were not identified through genome-wide association studies (GWASs) of the outcome. Thus, studies of heritable phenotypes can uncover biological pathways that will lead to a better understanding of basic mechanisms of disease and may identify targets for intervention. In a similar paradigm, mammographic density is a highly heritable, reliably measured, quantitative trait and a well-established strong predictor of breast cancer independent of known breast cancer risk factors. Identifying genes associated with mammographic density will identify mechanisms related to not only breast density, but has immense potential to detect genes involved with breast cancer. We propose to conduct a multi-stage GWAS of mammographic density among postmenopausal women (Aim 1). As part of the Cancer Genetic Markers of Susceptibility (CGEMS) project, postmenopausal breast cancer cases and controls in the Nurses'Health Study (NHS) have whole genome scans completed. We estimate that we will have mammographic density data on 1,800 of these women. Our initial analysis will examine the association between 2.5 million SNPs (includes 550,000 genotyped and the remainder imputed) and mammographic density among women included in the CGEMS project (Stage 1). To minimize false positive and negative associations, we will pursue the highest-ranking 7,600 SNPs from Stage 1 in an additional 1,200 postmenopausal women from the NHS (Stage 2). The 1,536 most promising SNPs will be genotyped in 3,000 postmenopausal participants in the Mayo Mammography Health Study (MMHS) (Stage 3).Validated SNPs that emerge from the multi-stage study will be evaluated for biologically plausible gene-environment interactions (Aim 2). The NHS and MMHS are both well established cohorts of demographically similar populations with blood samples, mammographic density data and extensive exposure information on breast cancer risk factors. We will also evaluate if validated SNPs from Aim 1 are associated with breast cancer risk in the NHS and in the Breast and Prostate Cancer Cohort Consortium (with over 6,000 breast cancer cases and controls). The results of the proposed study will complement those from breast cancer GWASs by increasing our understanding of breast biology and etiology of breast cancer. This is a unique, cost-efficient, and timely proposal to identify novel genetic pathways underlying breast density and breast cancer. Identification of genes associated with mammographic density will allow for study of their function as it relates to density and breast cancer and opens up the possibility for novel targets of breast cancer prevention and treatment. PUBLIC HEALTH RELEVANCE: Elucidating the genetic components of complex diseases with multifactorial causes such as breast cancer can be enhanced through concentration on heritable risk factors for the disease. Mammographic density is a highly heritable, reliably measured, quantitative trait and a well- established strong predictor of breast cancer independent of known breast cancer risk factors. This multi-stage genome-wide association study of mammographic density will not only identify novel loci associated with breast density, but will complement the studies of breast cancer by increasing our understanding of breast biology and etiology of breast cancer.